Signal recognition and background suppression by matched filters and neural networks for Tunka-Rex. (arXiv:1812.03347v1 [astro-ph.IM])

Signal recognition and background suppression by matched filters and neural networks for Tunka-Rex. (arXiv:1812.03347v1 [astro-ph.IM])
<a href="http://arxiv.org/find/astro-ph/1/au:+Shipilov_D/0/1/0/all/0/1">D. Shipilov</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Bezyazeekov_P/0/1/0/all/0/1">P. A. Bezyazeekov</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Budnev_N/0/1/0/all/0/1">N. M. Budnev</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Chernykh_D/0/1/0/all/0/1">D. Chernykh</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Fedorov_O/0/1/0/all/0/1">O. Fedorov</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Gress_O/0/1/0/all/0/1">O. A. Gress</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Haungs_A/0/1/0/all/0/1">A. Haungs</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Hiller_R/0/1/0/all/0/1">R. Hiller</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Huege_T/0/1/0/all/0/1">T. Huege</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Kazarina_Y/0/1/0/all/0/1">Y. Kazarina</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Kleifges_M/0/1/0/all/0/1">M. Kleifges</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Korosteleva_E/0/1/0/all/0/1">E. E. Korosteleva</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Kostunin_D/0/1/0/all/0/1">D. Kostunin</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Kuzmichev_L/0/1/0/all/0/1">L. A. Kuzmichev</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Lenok_V/0/1/0/all/0/1">V. Lenok</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Lubsandorzhiev_N/0/1/0/all/0/1">N. Lubsandorzhiev</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Marshalkina_T/0/1/0/all/0/1">T. Marshalkina</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Monkhoev_R/0/1/0/all/0/1">R. Monkhoev</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Osipova_E/0/1/0/all/0/1">E. Osipova</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Pakhorukov_A/0/1/0/all/0/1">A. Pakhorukov</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Pankov_L/0/1/0/all/0/1">L. Pankov</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Prosin_V/0/1/0/all/0/1">V. V. Prosin</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Schroder_F/0/1/0/all/0/1">F. G. Schröder</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Zagorodnikov_A/0/1/0/all/0/1">A. Zagorodnikov</a>

The Tunka Radio Extension (Tunka-Rex) is a digital antenna array, which
measures the radio emission of the cosmic-ray air-showers in the frequency band
of 30-80 MHz. Tunka-Rex is co-located with TAIGA experiment in Siberia and
consists of 63 antennas, 57 of them are in a densely instrumented area of about
1 kmtextsuperscript{2}. In the present work we discuss the improvements of the
signal reconstruction applied for the Tunka-Rex. At the first stage we
implemented matched filtering using averaged signals as template. The
simulation study has shown that matched filtering allows one to decrease the
threshold of signal detection and increase its purity. However, the maximum
performance of matched filtering is achievable only in case of white noise,
while in reality the noise is not fully random due to different reasons. To
recognize hidden features of the noise and treat them, we decided to use
convolutional neural network with autoencoder architecture. Taking the recorded
trace as an input, the autoencoder returns denoised trace, i.e. removes all
signal-unrelated amplitudes. We present the comparison between standard method
of signal reconstruction, matched filtering and autoencoder, and discuss the
prospects of application of neural networks for lowering the threshold of
digital antenna arrays for cosmic-ray detection.

http://arxiv.org/icons/sfx.gif